Fractal Print Pro - any good???
Keith Clark
They seem to be saying that you can get a scan from a desktop film
scanner (the review I read mentioned a Nikon LS2000), and "enlarge" the
scan to get a photographic quality 20x30 print.
Needless to say I'm a little doubtful.
I installed the demo version of their plug in, and used it to "enlarge"
a scan by 800%, then used Photoshop's bicubic interpolation to resample
the scan up 800%. It was really hard to see much difference on the
monitor.
Is this $300 plug-in really that much better than using Photoshop's
interpolation?
BTW, as a photographer, I know that resampling a scan doesn't add
information. I know you can't add quality data not present in the
original scan. It's just that I read a review by a big name wildlife
shooter, so I had to try it...and now I honestly can't see what the fuss
is all about.
Am I missing something here?
Feel free to reply via email.
Thanks for any info/opinions (good or bad).
Keith
Banshee
loss. You can enlarge as big as yuou want without jagged edges.
On 23 Nov 1999 05:46:12 GMT, Keith Clark
Keith Clark
Banshee wrote:
>
> Well a small PS file can be used to cover a building without data
> loss. You can enlarge as big as yuou want without jagged edges.
>
>
So are you saying you've used it professionally yourself, and do you
feel that it works better than Photoshop's bicubic interpolation?
Thanks,
Keith
Carl Schofield
images scaled up with genuine fractals plug in and photoshop's bicubic
interpolation.
----------
In article <383A29F8...@spiritone.com>, Keith Clark
Keith Clark
Someone said it was Moose Peterson who wrote the review. I wrote him and
the editors of the magazine the review appeared in, asking him basically
what I asked in my original post. Except I tried extra hard to be nice
and non-confrontational and explained that all I was after was his
opinion as to whether this expensive plug-in is really better than using
"image - adjust - image size (bicubic interpolation)" command and
running an edge sharpening afterwards, since those commands are free if
you have Photoshop already (Fractal Print Pro requires Photoshop).
Mr Peterson wrote back, and his response was rude, evasive and hostile.
I'm glad I didn't rush out and buy the software...
Keith
Jan Steinman -- jan AT bytesmiths DOT com [remove .gov]
ClarkPho...@spiritone.com wrote:
> Has anyone here got results with this program that justify it's price?
I have GF 2.0 -- cheaper than Print Pro, but otherwise, I don't know what
the differences are.
I'm moderately pleased with it. The artifacts it produces are more
"natural" looking than bicubic interpolation, IMHO.
> I installed the demo version of their plug in, and used it to "enlarge"
> a scan by 800%, then used Photoshop's bicubic interpolation to resample
> the scan up 800%. It was really hard to see much difference on the
> monitor.
Take a small section of the blown-up image, and look at it at 400% or so.
The artifacts for bicubic look "blotchy," "pixellated," or
"computer-like," in the words of a half-dozen friends I tested this out
with. The GF blow-up tends to make long artifacts where there are lines,
round artifacts where there are circles, etc. My test subjects described
it as "fuzzy," "warm," "abstract," or "painterly."
I've even used it at extremes to make abstract art from photographs. I
took a very small section of a cave photo (perhaps 100 pixels square) and
blew it up about 20x in GF. The stalactites and stalagmites looked
surreal, as though painted. Doing the same thing with Photoshop's
interpolation yielded something that looked, well, like it had been
extremely upsampled. :-)
> Is this $300 plug-in really that much better than using Photoshop's
> interpolation?
Only you can decide that. I got the cheaper version with a student
discount for about $150, and feel it was a reasonable investment. I've
made beautiful 16x32 prints from it, but you'd need a loupe to tell them
from a bicubic example.
> BTW, as a photographer, I know that resampling a scan doesn't add
> information. I know you can't add quality data not present in the
> original scan.
That's right! I'm pretty disappointed at GF's marketing. Did you see the
photo on the box? Fraudulent! It clearly "shows" GF making up detail where
none existed in the original scan! However, marketeers will be marketeers
-- I still believe GF has a place.
Summary: there's no such thing as free pixels, but GF seems to have a
slight edge over bicubic interpolation for producing "natural" looking
up-sampling artifacts.
--
: Jan Steinman -- Jan AT Bytesmiths DOT com
: Bytesmiths -- digital artistry <http://www.bytesmiths.com/Art_Gallery>
: +1 503 635 3229
Keith Clark
Thanks for the detailed analysis. Yours is the best I've read so far.
I think your comment that at 16x32 you needed a loupe to tell the prints apart
sums it up for me. (Although, you didn't say how big in MB the original scan
was...)
Nobody I know looks at mounted and framed prints with a loupe. So if one needs
a loupe to tell the difference between an upsampled scan using the "free"
tools built in to Photoshop, from a very expensive plug-in, I won't buy GF. I
too can get it at a substantial discount, but In my mind it's not worth any
amount of money.
Thanks again for your input! :>
Keith
Nicholas O. Lindan
>
> Nobody I know looks at mounted and framed prints with a loupe.
Ha!
Glad to meet you, Keith. I'm Nick Lindan, and I do look at mounted
and framed prints with a loupe. I have even taken a 20x loupe with
me to the museum.
David Hay Jones
> Glad to meet you, Keith. I'm Nick Lindan, and I do look at mounted
> and framed prints with a loupe. I have even taken a 20x loupe with
> me to the museum.
Is sharpness an end in itself? When we put pictures under the microscope
in this way the search for even the slightest amount of unsharpness
makes photography more about eliminating technical "flaws" than
creating lasting images. Ok, so creativity is not equal to blur, but nor
is good photography necessarily the same thing as razorsharpness.
I'd go so far as to say sharpness has become a fetish,
David
Keith Clark
I stand corrected. :>
Keith
"Nicholas O. Lindan" wrote:
> Keith Clark wrote:
> >
> > Nobody I know looks at mounted and framed prints with a loupe.
>
> Ha!
Jan Steinman -- jan AT bytesmiths DOT com [remove .gov]
ClarkPho...@spiritone.com wrote:
> Thanks for the detailed analysis. Yours is the best I've read so far.
Glad I could help!
> I think your comment that at 16x32 you needed a loupe to tell the prints apart
> sums it up for me. (Although, you didn't say how big in MB the original scan
> was...)
It was oversized to begin with. I used PhotoVista to stitch four frames
together. I believe that got me to about 200dpi at the target size. I then
used GF to go on up to 300dpi.
I agree that this is not a good use of GF, and I wouldn't have invested in
it simply to get 50% more pixels. In my prior posting, I tried to explain
where it might find use -- at expansions of >>200% or so.
There are a lot of other factors here, so I wouldn't judge it just by that
one statement. However, there is no free lunch or pixels!
Keith Clark
I finally did see a side by side comparison in the Design Graphics
article that someone pointed out. It was possible, but very difficult to
tell any difference in the extreme magnifications shown in the article.
If they (GF) would reduce the price to say $50-75 then I could see
-maybe- being able to justify the price. I'll certainly not pay the
retail price nor the discount price ($214).
Keith
"Jan Steinman -- jan AT bytesmiths DOT com [remove .gov]" wrote:
>